Fresh install of Fedora Core 3 and SSH crashes on startup with the following error (snipped from /var/log/secure): Dec 1 05:47:53 slave1 sshd[2390]: Server listening on :: port 22. Dec 1 05:47:53 slave1 sshd[2390]: error: Bind to port 22 on 0.0.0.0 failed: Address already in use.
What the heck is going on here and how do I fix it? I use my linux box through an SSH-tunneled VNC session from work for the most part because my wife is jealous of my computer. ;-)
Seriously, I need to be able to ssh into my box and work... any ideas? Thanks.
John,
John Aldrich wrote:
Fresh install of Fedora Core 3 and SSH crashes on startup with the following error (snipped from /var/log/secure): Dec 1 05:47:53 slave1 sshd[2390]: Server listening on :: port 22. Dec 1 05:47:53 slave1 sshd[2390]: error: Bind to port 22 on 0.0.0.0 failed: Address already in use.
What the heck is going on here and how do I fix it? I use my linux box through an SSH-tunneled VNC session from work for the most part because my wife is jealous of my computer. ;-)
Seriously, I need to be able to ssh into my box and work... any ideas? Thanks.
Seems like you already have sshd running (as I have the same message and no probs logging into my box). What does
ps axuwww | grep -i sshd
say? If the server crashes on startup, what is the message when you try to start it with
service sshd start
Alex
Alexander Apprich wrote:
John,
John Aldrich wrote:
Fresh install of Fedora Core 3 and SSH crashes on startup with the following error (snipped from /var/log/secure): Dec 1 05:47:53 slave1 sshd[2390]: Server listening on :: port 22. Dec 1 05:47:53 slave1 sshd[2390]: error: Bind to port 22 on 0.0.0.0 failed: Address already in use.
Seems like you already have sshd running (as I have the same message and no probs logging into my box). What does
This is the same question and same answer as encountered on FC2 with the 2.6 kernel because support was added for ipv6.
sshd is starting up and listening on all ipv6 addresses, then goes to listen on all ipv4 addresses and finds it is already doing so because the ipv6 scope includes the ipv4 scope. This is not a problem. If you do not need ipv6 you may disable it and avoid other potential pitfalls by adding the following line to /etc/modprobe.conf: alias net-pf-10 off
And then rebooting. Yes, this is a rare Linux scenario where a reboot is necessary to make a change.
If you do need ipv6 then just ignore the message.
Chris
On Wednesday 01 December 2004 08:07 am, Christopher K. Johnson wrote:
Alexander Apprich wrote:
This is the same question and same answer as encountered on FC2 with the 2.6 kernel because support was added for ipv6.
sshd is starting up and listening on all ipv6 addresses, then goes to listen on all ipv4 addresses and finds it is already doing so because the ipv6 scope includes the ipv4 scope. This is not a problem. If you do not need ipv6 you may disable it and avoid other potential pitfalls by adding the following line to /etc/modprobe.conf: alias net-pf-10 off
And then rebooting. Yes, this is a rare Linux scenario where a reboot is necessary to make a change.
If you do need ipv6 then just ignore the message.
It doesn't freakin' work! :( I've done what you suggested, and I can't ssh to my box from outside and I still get the message in the log file... I also tried "alias net-pf-6 off" as suggested in an archived message on the RedHat list.... *sigh* I /REALLY/ , really, wish that RedHat/Fedora wouldn't try to "make things easier" by installing IPV6 by default! It seems to create more headaches than it solves!
Can someone PLEASE help me turn off this IPV6 crap and get SSHD working? please????
John Aldrich wrote:
It doesn't freakin' work!
[snipped]
Can someone PLEASE help me turn off this IPV6 crap and get SSHD working? please????
Have you checked your firewall settings? It really does sound like a firewalling problem. Although you may have tried that already, I don't see it mentioned anywhere. Usually if it's a firewall problem, you'll get a connection refused error though. Depends on the client you are using for sshing into the box. What client -are- you using anyhow?
HTH
Alex White
Am Do, den 02.12.2004 schrieb Alex White um 4:58:
It doesn't freakin' work!
[snipped]
Can someone PLEASE help me turn off this IPV6 crap and get SSHD working? please????
Have you checked your firewall settings? It really does sound like a firewalling problem. Although you may have tried that already, I don't see it mentioned anywhere. Usually if it's a firewall problem, you'll get a connection refused error though. Depends on the client you are using for sshing into the box. What client -are- you using anyhow?
Alex White
Yes, I bet the the same and answered in this direction to the other thread he opened hours ago.
netstat -talpen | grep ":22" | grep LISTEN
Will very certainly show the SSH daemon perfectly listen on port 22. He wrote:
"Just to restate the problem -- when I start my linux box I do NOT manually attempt to start sshd. If I attempt to ssh into my linux box, I get a timeout error. "
A timeout indicates a package DROP. If the daemon would not listen a reject would appear. So it is very certain a wrong firewall setup.
Alexander
Alexander Dalloz wrote:
Am Do, den 02.12.2004 schrieb Alex White um 4:58:
[snipped]
Have you checked your firewall settings? It really does sound like a firewalling problem.
[snipped]
Yes, I bet the the same and answered in this direction to the other thread he opened hours ago.
A timeout indicates a package DROP. If the daemon would not listen a reject would appear. So it is very certain a wrong firewall setup.
Alexander
Oh, okay got ya, I missed that part of the thread, as I was fixing my own idiotic shorewall issue with postfix. =-(
On Wednesday 01 December 2004 10:58 pm, Alex White wrote:
Have you checked your firewall settings? It really does sound like a firewalling problem. Although you may have tried that already, I don't see it mentioned anywhere. Usually if it's a firewall problem, you'll get a connection refused error though. Depends on the client you are using for sshing into the box. What client -are- you using anyhow?
Yeah... that's it all right. The guy who runs the local LUG here told me how to disable iptables to test and once I did that, I was able to get right in. Now, all I need to know is how to enable SSH and still have the rest of the iptables running! I'm afraid I never learned how to write iptables (or ipchains for that matter) rules. Thanks John
John Aldrich wrote:
On Wednesday 01 December 2004 10:58 pm, Alex White wrote:
Have you checked your firewall settings? It really does sound like a firewalling problem. Although you may have tried that already, I don't see it mentioned anywhere. Usually if it's a firewall problem, you'll get a connection refused error though. Depends on the client you are using for sshing into the box. What client -are- you using anyhow?
Yeah... that's it all right. The guy who runs the local LUG here told me how to disable iptables to test and once I did that, I was able to get right in. Now, all I need to know is how to enable SSH and still have the rest of the iptables running! I'm afraid I never learned how to write iptables (or ipchains for that matter) rules. Thanks John
If you emacs /etc/sysconfig/iptables, you will see some various input. From here you can add the following:
-A RH-Firewall-1-INPUT -m state --state NEW -m tcp -p tcp --dport 22 -j ACCEPT
There is going to be some line wrapping issues with that, just type it all on one line, (Don't hit return). The line will wrap automatically depending on how big the window is for emacs. Of course use the editor of your choice. As a sample of what my /etc/sysconfig/iptables lookes like, here is a section:
-A RH-Firewall-1-INPUT -m state --state ESTABLISHED,RELATED -j ACCEPT -A RH-Firewall-1-INPUT -m state --state NEW -m tcp -p tcp --dport 22 -j ACCEPT -A RH-Firewall-1-INPUT -m state --state NEW -m tcp -p tcp --dport 21 -j ACCEPT
Note that the ssh and ftp port in my iptables setup comes -after- the rule which allows already established connections out from my machine back in. (Alexander can expound upon this if you don't know what the duece I'm talking about, or I will depending on which one of us gets the question first hehe).
HTH
Alex White
Quoting Alexander Apprich A.Apprich@science-computing.de:
Seems like you already have sshd running (as I have the same message and no probs logging into my box). What does
ps axuwww | grep -i sshd
say? If the server crashes on startup, what is the message when you try to start it with
service sshd start
Well, I never get any error message on the screen. What I am doing is rebooting my box and then logging in. I am not manually starting sshd. I've tried "service restart sshd" and I get the same error message in the log. I'll check the ps you suggest when I get home...